Narrative:

This incident was caused by new 'descend via' procedures associated with RNAV stars at atl. While working a CRJ2 on the HONIE5 arrival; I instructed the aircraft to cross fogog at 12000 ft and then descend via the honie 5 arrival. I then observed the aircraft starting a descent from 12000 ft through 11700 ft approximately 7 mi prior to fogog. I advised the aircraft to maintain 11000 ft which was still in my airspace and safe from potential conflicting traffic at lower altitudes in that area. When the aircraft initially entered the airspace (approximately 40 NM southeast of atl) I had deleted their speed restrs and descended them from 14000 ft to 12000 ft. By issuing the aircraft control instructions contrary to the STAR; I forced the pilots to amend the flight plan in their computer. The head down time required for this probably contributed to the error. Atlanta approach (A80) has just begun using RNAV stars for atl which include speed restrs and altitude crossing restrs. There have been numerous (I believe more than 10) incidents similar to the one described here since the stars became effective. It is only due to the diligence of ATC controllers that have prevented conflicts or worse. The problem seems to be a combination of pilot education; complexity of the procedures; air traffic volume; and the aircraft's FMC. Pilot education is probably the most crucial item; however; it may be the hardest to accomplish. The pilots based at atl will probably become proficient at the procedures in the near future. The more dangerous problem is there will always be pilots coming in that are not familiar with the airport. As local controllers and pilots become comfortable with the procedures they will also become more complacent. When an unfamiliar pilot comes in and makes an error it could lead to a disaster if it is not recognized by the controller. The procedures are extremely complicated. In addition to the multiple speed and altitude restrs; there are numerous potential runway transitions. The same chart covers different landing directions (east versus west) with completely different expectations. Due to traffic volume at atl; aircraft are frequently issued alternate instructions during their arrival for sequencing purposes. This may be speed adjustments and altitude assignments as well as vectors off the procedure. The expected landing runway is also frequently changed as our flow controllers are trying to balance demand on 3 arrival runways. In addition; pilots arriving to atl are often tasked with additional duties preparing for triple ILS RPM approachs. The FMC; like any computer; is a wonderful tool for pilots when given proper input. However; given the intense situation the pilots face arriving into atl; I don't think it is practical to expect them to be making multiple changes to their flight plan during this phase of flight. In an ideal world an aircraft would be cleared on the appropriate STAR by the en route center. Upon checking in with the approach controller they would be assigned the appropriate landing runway transition and instructed to descend via STAR. In that scenario the computer would probably have all the appropriate information to safely fly the route. This is not a practical reality for an airport as busy as atl. I think we need to keep it as simple as possible. Just a few yrs ago we vectored all arrs to final and all departures out of the airspace. By doing this; all the pilots had to worry about from ATC was heading; speed; and altitude. We had 4 runways and atl was by far the most efficient airport in the world. We have now added a 5TH runway and RNAV arrs and departures. While there may be improvements in some areas; safety has been compromised. An airport this busy is inherently complex and we are making it worse with too many changes in a relatively short period of time. RNAV arrs have some benefits; however; descend via procedures are too complex; unsafe; and offer very little benefit from an ATC standpoint.

Google
 

Original NASA ASRS Text

Title: A80 CTLR DESCRIBED AN ACR EARLY DESCENT EVENT; POSSIBLY LINKED TO 'DESCEND VIA' PHRASEOLOGY AND NEWLY INITIATED RNAV ARR PROCS.

Narrative: THIS INCIDENT WAS CAUSED BY NEW 'DSND VIA' PROCS ASSOCIATED WITH RNAV STARS AT ATL. WHILE WORKING A CRJ2 ON THE HONIE5 ARR; I INSTRUCTED THE ACFT TO CROSS FOGOG AT 12000 FT AND THEN DSND VIA THE HONIE 5 ARR. I THEN OBSERVED THE ACFT STARTING A DSCNT FROM 12000 FT THROUGH 11700 FT APPROX 7 MI PRIOR TO FOGOG. I ADVISED THE ACFT TO MAINTAIN 11000 FT WHICH WAS STILL IN MY AIRSPACE AND SAFE FROM POTENTIAL CONFLICTING TFC AT LOWER ALTS IN THAT AREA. WHEN THE ACFT INITIALLY ENTERED THE AIRSPACE (APPROX 40 NM SE OF ATL) I HAD DELETED THEIR SPD RESTRS AND DSNDED THEM FROM 14000 FT TO 12000 FT. BY ISSUING THE ACFT CTL INSTRUCTIONS CONTRARY TO THE STAR; I FORCED THE PLTS TO AMEND THE FLT PLAN IN THEIR COMPUTER. THE HEAD DOWN TIME REQUIRED FOR THIS PROBABLY CONTRIBUTED TO THE ERROR. ATLANTA APCH (A80) HAS JUST BEGUN USING RNAV STARS FOR ATL WHICH INCLUDE SPD RESTRS AND ALT XING RESTRS. THERE HAVE BEEN NUMEROUS (I BELIEVE MORE THAN 10) INCIDENTS SIMILAR TO THE ONE DESCRIBED HERE SINCE THE STARS BECAME EFFECTIVE. IT IS ONLY DUE TO THE DILIGENCE OF ATC CTLRS THAT HAVE PREVENTED CONFLICTS OR WORSE. THE PROB SEEMS TO BE A COMBINATION OF PLT EDUCATION; COMPLEXITY OF THE PROCS; AIR TFC VOLUME; AND THE ACFT'S FMC. PLT EDUCATION IS PROBABLY THE MOST CRUCIAL ITEM; HOWEVER; IT MAY BE THE HARDEST TO ACCOMPLISH. THE PLTS BASED AT ATL WILL PROBABLY BECOME PROFICIENT AT THE PROCS IN THE NEAR FUTURE. THE MORE DANGEROUS PROB IS THERE WILL ALWAYS BE PLTS COMING IN THAT ARE NOT FAMILIAR WITH THE ARPT. AS LCL CTLRS AND PLTS BECOME COMFORTABLE WITH THE PROCS THEY WILL ALSO BECOME MORE COMPLACENT. WHEN AN UNFAMILIAR PLT COMES IN AND MAKES AN ERROR IT COULD LEAD TO A DISASTER IF IT IS NOT RECOGNIZED BY THE CTLR. THE PROCS ARE EXTREMELY COMPLICATED. IN ADDITION TO THE MULTIPLE SPD AND ALT RESTRS; THERE ARE NUMEROUS POTENTIAL RWY TRANSITIONS. THE SAME CHART COVERS DIFFERENT LNDG DIRECTIONS (E VERSUS W) WITH COMPLETELY DIFFERENT EXPECTATIONS. DUE TO TFC VOLUME AT ATL; ACFT ARE FREQUENTLY ISSUED ALTERNATE INSTRUCTIONS DURING THEIR ARR FOR SEQUENCING PURPOSES. THIS MAY BE SPD ADJUSTMENTS AND ALT ASSIGNMENTS AS WELL AS VECTORS OFF THE PROC. THE EXPECTED LNDG RWY IS ALSO FREQUENTLY CHANGED AS OUR FLOW CTLRS ARE TRYING TO BAL DEMAND ON 3 ARR RWYS. IN ADDITION; PLTS ARRIVING TO ATL ARE OFTEN TASKED WITH ADDITIONAL DUTIES PREPARING FOR TRIPLE ILS RPM APCHS. THE FMC; LIKE ANY COMPUTER; IS A WONDERFUL TOOL FOR PLTS WHEN GIVEN PROPER INPUT. HOWEVER; GIVEN THE INTENSE SITUATION THE PLTS FACE ARRIVING INTO ATL; I DON'T THINK IT IS PRACTICAL TO EXPECT THEM TO BE MAKING MULTIPLE CHANGES TO THEIR FLT PLAN DURING THIS PHASE OF FLT. IN AN IDEAL WORLD AN ACFT WOULD BE CLRED ON THE APPROPRIATE STAR BY THE ENRTE CTR. UPON CHKING IN WITH THE APCH CTLR THEY WOULD BE ASSIGNED THE APPROPRIATE LNDG RWY TRANSITION AND INSTRUCTED TO DSND VIA STAR. IN THAT SCENARIO THE COMPUTER WOULD PROBABLY HAVE ALL THE APPROPRIATE INFO TO SAFELY FLY THE RTE. THIS IS NOT A PRACTICAL REALITY FOR AN ARPT AS BUSY AS ATL. I THINK WE NEED TO KEEP IT AS SIMPLE AS POSSIBLE. JUST A FEW YRS AGO WE VECTORED ALL ARRS TO FINAL AND ALL DEPS OUT OF THE AIRSPACE. BY DOING THIS; ALL THE PLTS HAD TO WORRY ABOUT FROM ATC WAS HDG; SPD; AND ALT. WE HAD 4 RWYS AND ATL WAS BY FAR THE MOST EFFICIENT ARPT IN THE WORLD. WE HAVE NOW ADDED A 5TH RWY AND RNAV ARRS AND DEPS. WHILE THERE MAY BE IMPROVEMENTS IN SOME AREAS; SAFETY HAS BEEN COMPROMISED. AN ARPT THIS BUSY IS INHERENTLY COMPLEX AND WE ARE MAKING IT WORSE WITH TOO MANY CHANGES IN A RELATIVELY SHORT PERIOD OF TIME. RNAV ARRS HAVE SOME BENEFITS; HOWEVER; DSND VIA PROCS ARE TOO COMPLEX; UNSAFE; AND OFFER VERY LITTLE BENEFIT FROM AN ATC STANDPOINT.

Data retrieved from NASA's ASRS site as of May 2009 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.